How corporate giants are automating the farm
In 2012 agribusiness giant Monsanto purchased an obscure data startup for an amount just short of a cool one billion dollars. Its new acquisition, The Climate Corporation, was a digital provider of weather and climate information to crop insurance companies. As far as anyone could tell, it had nothing to do with genetically engineering seeds, mixing pesticides or the other activities for which Monsanto is notorious. Many observers were left scratching their heads: a billion dollars for a weather forecaster? – what on earth was afoot?
The answer may lie in the biggest technological shifts in industrial agriculture since the end of World War Two. Today Monsanto claims they are not so much a seed company as a data company. The Climate Corporation now stands at the centre of a fundamental reorganization not just at Monsanto but across the entire agribusiness sector, precipitating a wave of mega-deals. As Monsanto merges with Bayer, DuPont with Dow Chemical, and Syngenta with ChemChina, the $400 billion agricultural inputs market is suddenly crystalizing into a highly concentrated three-way monopoly in anticipation of a further merger wave. This new cartel sees neither seed, soil amendments nor sprays as its key product but trucks instead in big data, coupled to artificial intelligence (AI) and new automation capabilities that are reshaping farming and the food system.
Pixel by pixel
This new wave goes by the name ‘precision agriculture’ (or as Monsanto prefer: ‘digital agriculture’). The vision is one of extreme mechanization down on the farm, driven by algorithmic processing of big data. The task of raising food among the elements is the perfect sort of ‘optimization problem’ to which tech companies are now applying AI tools to score Silicon Valley investment. Records of weather, soil moisture, pests and crop history can all be crunched as datasets and then run through machine-learning algorithms to inform automated farming machinery. In glossy presentations of precision agriculture, the modern ‘data-driven’ farmer tromps through the fields staring not at the corn but at an iPad.
Meanwhile, self-driving tractors and agricultural robots co-ordinate with data from satellites, sensors and scouting drones that provide the farmer with real-time information about every square centimetre of earth to propose what seed, fertilizer, fungicide or pesticide ought to be applied exactly where and exactly when to maximize yield. Fields on The Climate Corporation’s ‘farmerview’ software platform are rendered as colourful pixelated geolocation maps, screaming in bright colours which corner of which field needs to be drilled or sprayed. Monsanto claims to have maps and historic data on all of the 30 million agricultural fields in the US – and that’s just a start.
It’s not clear that gene giants such as Monsanto will end up as the kings of this new automated farmscape. The current mega-mergers among big input companies are in part a defensive move to fend of challenges by traditional agricultural equipment makers to dominate this space – including names such as John Deere, AGCO and Case New Holland (CNH). Over the past few years, they have gone on a dizzying spree of acquiring data, sensors, drones and robotic partnerships.
Near daily announcements herald this new robotic lettuce picker or that self-driving combine harvester. Since at least 2001, the world’s largest tractor-maker John Deere has had satellite-connected GPS with centimetre-scale accuracy embedded into every new tractor. More recently, it has placed sensors in the nozzles, hoes and fertilizer applicators so that every pass through the field is also a data-gathering exercise tracking yield, weather and soil conditions to drive the computer models.
Giving more power to John Deere and Monsanto to determine how the farm is managed moves in the opposite direction from that called for by the ‘food sovereignty’ movement led by small farmers. It also brings the food movement into new disputes. Advocates claim that precision farming, with its targeted use of pesticides, will bring about a dramatic decrease in the use of chemicals in industrial agriculture (they said the same about GMOs), even being able to meet the requirements for organic certification.
But while the use of robots to remove weeds may provide an easy switch for industrial farmers who want to break the addiction to chemical weedkillers, letting algorithms decide every move means giving up on farmers’ knowledge of seed, soil and technique, while trusting the algorithms not to be biased in Monsanto’s commercial favour. It strays further into a future of people-less fields where only drones and robots know the land – or think they do. Organic farmers warn that farming organic by robot risks losing the benefits of agro-ecological production in which a farmer builds up soil and eco-system health over many years. A precision-managed system may be ‘just about’ organic enough to pass certification standards but have no deeper resilience when the unexpected hits.
Much investment is going into automating the farming of high-value crops such as berries, vegetables, fruits and spices that currently rely on labour – either by small-scale producers or low paid, migrant workers. A recent presentation by tech analysts Lux Research highlighted how Driscoll’s, the US berry giant, could potentially use one agricultural robot to analyse the colour and ripeness of berries and then selectively pick them, and replace up to 10 workers at a time. That has a human cost. In the US, Driscoll’s berry pickers are typically vulnerable Mexican-American migrant workers. Farm workers in both the US and Mexico already launched a boycott of Driscoll’s in 2015 over their failure to uphold workers’ rights, fair pay and healthy conditions. ‘If the intent is to bring machines into the field, that’s upsetting,’ Erik Nicholson, national vice-president for United Farm Workers recently told Bloomberg. ‘I want our food produced by people who are knowledgeable and able to earn a fair living.’
This may also explain why the current Trump administration feels so sanguine about plans to build a wall and clamp down on immigration, shrinking the labour pool that currently sustains US agriculture. American robots, so the logic goes, can replace Mexican farmworkers.
Livestock too is getting automated. From shepherding by drone (yes, really) to fitting out animals with sensors to monitor health remotely, meat and dairy producers are pursuing the next level of agricultural robotics. Fujitsu’s connected cow project, fits dairy cows with 3G or 5G connected wireless devices so that a farmer can track the readiness of specific cows for artificial insemination. Meanwhile, the so-called ‘internet of cows’, a data-tracking system from Brazil, assigns every cow its own webpage with basic data to be tracked by cellphone app. And in factory farms, body-borne sensors track growth, reconceptualizing animals as data to be optimized.
For now, the ag-giants mostly have large-scale Northern farmers as customers. But this is just the first wave. Further down the line, small farmers too are firmly in the crosshairs of precision agriculture’s coming disruption. These new tools may allow industrial monoculture operations to move onto land previously considered uneconomic – the so called ‘marginal lands’ where smallholder peasants still produce 70 per cent of the food that feeds the world. Players such as the Bill & Melinda Gates Foundation are actively exploring the potential for precision agriculture to upgrade these peasant farms to embrace more mechanized, data-driven farming models. That may in time prove to be a mistake. If a drone can map it and a robot can farm it, why would an ag corporation not move the peasants off the land, seize their soil and bring in the agbots – massively extending the global land grab one data-driven, precision-farmed square centimetre at a time.
Jim Thomas is Programme Director of ETC Group: etc.group.org